There is a move that used to belong to the very rich and now belongs to anyone with a brokerage account and a phone. You do not sell the stock. You borrow against it. The bank calls it a securities-backed line, a pledged-asset line, a margin loan; the pitch is the same under every name. Keep the position, keep the growth, pay no capital-gains tax, and have the money anyway.
It is a genuinely useful instrument. It is also the easiest way in personal finance to feel richer while becoming poorer, because every part of it is designed to leave the portfolio looking untouched.
Ask someone with such a line what it is costing them and they will quote the rate. Ask what it has cost them, in dollars, against what the borrowed money earned, over the years it has been open, and the answer is a pause. Nothing they use was built to tell them.
Why the cost is invisible
Sell a share and something visible happens. The position shrinks, the cash appears, a tax form arrives in February. The transaction leaves a mark on every statement it touches.
Borrow against the share and nothing visible happens at all. The position sits where it was, at full value, in an account whose name might now say “pledged” in small type. The loan lives on a different statement from a different department. The interest is debited monthly from cash, often the cash in the collateral account itself, so that it never crosses the household’s checking account and is never seen as spending. The brokerage app, opened on a good day, shows the same green number it always did.
This is not deception. It is the ordinary result of each provider reporting the one thing it can see. The lending desk sees a loan. The brokerage sees a portfolio. Neither sees the household, which is the only place the two numbers meet.
Family offices meet them on one page as a matter of routine. The line’s balance sits under the portfolio it is secured by, its interest is a line item beside the portfolio’s return, and the two are compared every quarter under a heading like “cost of carry.” Not because borrowing is wrong. Because a borrowing whose cost nobody adds up is a decision nobody is making.
The arithmetic
Four steps, using illustrative round numbers. They show a method; they do not describe anyone, and none of it is advice.
Step one: the carry. A household borrows $1,000,000 against a $3,000,000 portfolio at 7%. The interest is $70,000 a year. That is the plain cost, and it is paid whether markets rise or fall. Whatever the borrowed million was used for, the portfolio it is secured by now has to earn $70,000 a year, on top of everything else it was going to do, merely for the household to stand still.
Step two: what selling would have cost instead. Suppose the household wanted $1,000,000 in hand and the stock is worth two and a half times what it cost, so that 60 cents of every dollar sold is gain. Capital-gains tax falls on the gain, not on the proceeds. To net a million the household sells about $1,167,000 of stock, of which $700,000 is gain; at a combined 23.8% federal rate the tax on that gain is about $167,000, and $1,000,000 is left in hand. (Had every dollar sold been gain, the tax would have been $238,000 on the same sale; the basis is what keeps it lower.) Borrowing avoids that payment now. This is the whole case for the line, and it is a real one: $167,000 kept invested is $167,000 that goes on compounding.
The arithmetic behind it:
Stock sold: $1,166,600
Gain in it, at 60 cents per dollar: $700,000
Tax at 23.8% of the gain: $166,600
Left in hand: $1,000,000
But notice what was actually bought. The tax was not cancelled. It was deferred, and the household paid $70,000 in year one for the deferral. In year two it pays another $70,000. By the middle of year three the interest paid exceeds the tax that was avoided, and the tax is still owed the day the stock is finally sold. The line does not remove the bill. It rents time.
Step three: whether the rent was worth it. Renting time is worth it when the thing you kept earns more than the rent. Compare two households. One sold $1,167,000 of stock and has $1,000,000 in hand, tax paid, no loan. The other borrowed, has the same $1,000,000 in hand, still owns the $1,167,000 of stock, and owes $1,000,000 at 7%.
The second household is ahead by exactly what the retained $1,167,000 earns, and behind by $70,000 a year. Break-even is a return of about 6% on the retained stock. Above that, borrowing was the better move; below it, selling was. In a year the portfolio returns 12%, the line earned its keep handsomely. In a year it returns 2%, the household paid $70,000 to keep an asset that made $23,000. In a year it falls 20%, the household lost $233,000 on stock it could have sold, and paid $70,000 for the privilege.
Written as a sentence: the line is a bet that the portfolio will beat its own borrowing rate, renewed every year, with the household’s other assets as the stake. Most people who hold one have never seen the sentence. They have seen the rate.
Step four: the mark. A pledged line has a rule inside it that a mortgage does not. The lender will advance only so much against the collateral, commonly half or less of its value, and the ratio is tested every day. With $3,000,000 securing $1,000,000, the household is borrowing a third of the collateral. If the portfolio falls 30%, the collateral is $2,100,000 and the loan is 48% of it, a whisker under the limit. If the portfolio falls 40%, the collateral is $1,800,000, the loan is 56% of it, and the lender asks for $100,000 by tomorrow, or sells stock to get it.
That is the part the arithmetic in step three leaves out. The line does not merely cost interest. It converts a market fall into a liquidity event, at the exact moment the household is least able to meet one, and the forced sale realizes the very gain the line was opened to defer, now at a lower price. The stock a household can hold through a crash is stock nobody can make it sell. Pledged stock is not that.
How it shows up in the wealth picture
Here is where the invisibility does its real damage, because it reaches the one number that is supposed to be honest.
Net worth is assets less debts. Borrow $1,000,000 and spend it, and net worth falls by whatever was spent, as it should. Borrow $1,000,000 and reinvest it, and net worth does not move at all: the new asset and the new debt cancel. The portfolio, though, is now $4,000,000 supported by $3,000,000 of the household’s own money, and every percentage point the market moves is worth a third more to the household than it was, up and down alike.
A dashboard that shows the portfolio’s growth without the line’s growth beside it is now reporting on a bigger portfolio than the household owns. In a rising year it flatters; the assets grew, the line stood still, the household looks like a better investor than it was. In a falling year it slanders, in the same proportion. The honest reading is three numbers together: how much the assets grew, how much the borrowing grew, and what the difference did to the household’s own equity. On the household from step one, whose own equity is $2,000,000, a year in which the $3,000,000 portfolio rose 8% while the line grew by $150,000 of interest and further draws is a year in which the assets made $240,000 and the household’s equity rose $90,000, about 4½%. Both numbers are true. Only one of them is the household’s.
The interest itself deserves a line of its own. Debited from the collateral account, it reads on a feed as a withdrawal, money the household took out. It is nothing of the kind: the household took nothing; the lender did. Counted as a withdrawal it disappears into the same bucket as a transfer to checking, and the cost of the line is once again invisible, this time inside the very tool that was supposed to add it up. It is a cost of carrying the position, and it belongs beside the position’s return, not among the household’s transfers.
What the number is not
It is not an argument against borrowing. The line is a tool, and for a household with a concentrated position, a large one-off need and a long horizon, it can be the right one. The failure mode is not borrowing. It is borrowing without the four numbers: the carry, the tax deferred, the break-even return, and the fall that triggers the call.
It is not a forecast. Whether the portfolio beats the rate next year is unknowable, and a tool that said otherwise would be selling advice. What can be known is what happened last year, and the year before, and whether the bet has been paying.
And it is not a monthly number. A carry is a slow cost against a slow return. Once a quarter, the line’s interest for the period beside the portfolio’s return for the period, is enough to see whether the arrangement is earning its keep.
The older maxim
Three sayings were carved at the entrance to the temple at Delphi. Two of them everyone knows: know thyself, and nothing in excess. The third is less quoted, and it is about exactly this. Engýa, pára d’áta: give a pledge, and ruin is near. The Greeks meant standing surety for another person’s debt, and every farmer who lost his land to a neighbor’s default understood why it belonged beside the other two.
A pledged line is standing surety for your own future self. The household of today borrows against the household of the year the market falls, and signs on its behalf. That can be a fair bargain. The Delphic point is only that it should be a knowing one, and the first maxim on the wall says how: know what you are pledging, what it costs to carry, and what happens to it on the bad day.
Measurement does not answer whether to borrow. It puts the pledge on the same page as the portfolio, which is where the decision was always being made, whether or not anyone was watching.
Nothing here is advice. It’s the arithmetic, explained plainly.
If you want the instrument behind the writing: Eudos is a family office in software. It shows a portfolio’s growth beside the growth of the borrowing secured on it, books the line’s interest as a cost of carry rather than money withdrawn, and keeps the household’s own equity as the number that matters, alongside every account, property, entity and collection in one encrypted place. It’s in private beta; the waitlist is at eudos.com. To see it first with a demo family's numbers, no sign-up: eudos.com/demo